1. Field of the Invention
This invention relates to a crystal growth process for producing large area group III-V compounds and particularly GaAs having controllable resistivity and the use thereof for fabrication of infrared window/domes with EMI/EMP protection.
2. Background and Brief Description of the Prior Art
All electro-optical (EO) systems can be easily affected by electromagnetic interference (EMI) and electro-magnetic pulses (EMP). The source of such interference can be of various types, including radio frequencies (RF), microwaves, magnetic fields, nuclear pulses, etc. emanating external of the electro-optical system. Such interference can adversely affect the operation of the EO system or, at times, destroy the system and particularly, its sensor. Since all EO systems usually have a window or dome separating them from the external environment, it is desirable and even often required that the window or dome provide EMI/EMP protection.
In the prior art such protection has generally been obtained by application of a conductive coating or grid, such as a metal window screen, to the window or dome. This conductive coating or grid couples with any EMI/EMP, thereby reflecting and, to a much lesser extent, absorbing this radiation. Thus, the EO system is shielded. There are, however, serious drawbacks to the use of conductive coatings and grids of windows and domes. These include transmission losses in the infrared region (the usual operational wavelengths of the EO system), modulation transfer function (MTF) or resolution losses due to diffraction effects and limited EMI/EMP protection of less than 30 dB at 1 GHz. It is generally essential that a substantial portion of the desired frequency spectrum be transmitted through the dome or window.
Gallium arsenide (GaAs) and gallium phosphide (GaP) are currently the only materials known that are optically transparent in the 1 to 14 micrometer (micron) wavelength region while their resistivity and temperature can vary over a wide range. However, other group III-V and group II-VI compounds should also have this property.
Ge is very temperature sensitive and the resistivity is too high to offer the same EMI protection as would GaAs. Silicon (Si) has been suggested for use as an infrared window as set forth in U.S. Pat. No. 4,826,266, however silicon is infrared transmissive only in the 3 to 5 micron range and not in the 8 to 12 micron range. The discussion hereinafter will be directed to GaAs, it being understood that GaP is also included as an appropriate material in accordance with the invention and that other group III-V and group II-VI compounds could possibly also be used. GaAs is, however, very expensive and difficult to fabricate with large dimensions on the order of the dimensions of optical system windows and domes. Furthermore, the uniform doping required to control and adjust the resistivity of GaAs has been very difficult if not impossible to achieve in the prior art.
In the prior art, there are two processes which have been generally used to produce optical quality GaAs, these being liquid encapsulated Czochralski (LEC) and Bridgman. The LEC method is limited in the size wafer it can produce. Bridgman techniques, on the other hand, have been used to produce large (8 inch.times.12 inch) GaAs windows but with non-uniform doping and at great expense.
The LEC method involves loading gallium into a crucible, usually made of boron nitride. Boron oxide glass is placed on top of the gallium. Arsenic is loaded into an injector cell, which is usually a hollow fused quartz vessel with a tube extending therefrom. The crucible containing the gallium and boron oxide glass and the arsenic-containing injector are placed within a heated pressure vessel and the crucible is heated above the melting point of GaAs. The boron oxide glass softens and flows over the gallium to encapsulate it. This glass encapsulant and the pressure in the vessel prevent liquid GaAs from decomposing. The arsenic injector is then positioned and heated so that the arsenic sublimes into the gallium forming GaAs liquid (some systems mix the gallium and arsenic in the crucible and compound them under higher pressures, thereby doing away with the injector). The injector is removed from the crucible and a seed crystal is dipped into the top of the liquid GaAs. The seed is then pulled upward such that a GaAs crystal grows therefrom. The LEC process requires several operations as well as manpower and is therefore expensive. Large size crystals cannot be grown due to the seed pulling operation and the required dimensions and temperature uniformity of the heater.
Bridgman growth involves the translation of the melt (liquid GaAs) through a temperature gradient that causes crystal growth. The furnace, melt or temperature gradient can be translated in any direction to accomplish the growth. To utilize the Bridgman process, the gallium and arsenic are loaded into opposite ends of a fused quartz ampoule which is then sealed. The gallium is heated above the melting point of GaAs and the arsenic is heated to a temperature such that it sublimes and reacts with the gallium, forming GaAs. The temperature gradient is then translated across the melt which freezes or grows the GaAs crystal. The sizes that can be produced in a Bridgman system are limited by the availability of a functional ampoule. The labor and material (the ampoule) also make the Bridgman process expensive. Since the normal growth direction is horizontal, this produces non-uniformities in the GaAs doping and therefore non-uniformities in the resistivity across the GaAs crystal. This is due to the natural segregation of the dopant during growth.
It is known that standard undoped, high resistivity semiconductor grade GaAs is highly transmissive to infrared frequencies in the range of 1 to 14 microns and is opaque from about 14 to 30 or 40 microns but then again becomes transparent. It is also known that the above described GaAs is opaque from about 1 micron downward to about the x-ray range, though the question of opacity below the 1 micron range is not of concern herein. However, such material is not useful as an EMI protective optical window or dome unless it can be made opaque to the entire spectrum above 14 microns. It is therefore apparent that an optical window or dome which is capable of transmitting the desired infrared frequencies (3 to 5 and 8 to 12 microns) with no or minimal loss, yet is opaque to frequencies below the useful infrared frequency range, is highly desirable. It is further apparent that a crystal growth process for producing optical grade group III-V compounds and particularly GaAs or GaP that overcomes the above noted limitations is also desirable.